Concentration from MAW-package in MODFLOW6

[TeunvanDooren]

Hi all,

Currently I am busy with a model that includes an extraction well in an aquifer with a fresh/saline transition zone. I am using the multi-aquifer well package (ModflowGwfmaw()) to simulate the extraction, and I want to follow the increasing salt concentration in the well continuously. I know I can get that data from the budget files (Mass_budget / Water_budget). However, given the size of the budget files, I want to avoid storing the full model budget for multiple timesteps.

I am using ModflowGwfmaw().obs.initialize to observe the head and extraction rate of the well. However, for transport, maw is included in ModflowGwtssm, which does not have the .obs option.

My model starts at r = r_well, i.e. my MAW well is outside my axisymmetric model grid. I already tried to get the concentration by observing the concentrations in the connected cells (see below). Is there also a direct way to observe the concentration in the MAW itself?

    # --- split well_con into per-well cell lists ---
    well1_connection_cells = []
    for rec in well_cod:
        wellid, _, cellid = rec[0], rec[1], rec[2]
        lay, row, col = cellid
        if wellid == 0:
            well1_connection_cells.append((lay, row, col))

    # --- build obs dictionary ---
    obs_file = f"{sim_name}.maw_conc{specie}.obs"
    csv_file = obs_file + ".csv"
    obs_dict_maw_conc = {csv_file: []}

    # Well 1 conc obs
    for i, (lay, row, col) in enumerate(well1_connection_cells):
        obs_dict_maw_conc[csv_file].append(
            (f"conc_w1_cell{i}", "concentration", (lay, row, col))
        )

    # --- create obs package for GWT ---
    flopy.mf6.ModflowUtlobs(
        gwt,
        pname=f"{specie}_obs",
        filename=obs_file,
        continuous=obs_dict_maw_conc,
        digits=10,
        print_input=True
    )

Teun

[mbakker7]

@TeunvanDooren, this was a good question, but I see you closed it. It would help me (and possibly future users) if you post a brief answer or solution here. So I reopen the discussion.

[TeunvanDooren]

I slightly edited my post. The code above does work, but maybe there is a more elegant/direct way to get the concentrations in the MAW-well.

[mbakker7]

But if I understand you correctly, the code that you provide does not give the same answer as the budget file. So the budget file computes it differently. Maybe somebody knows how the budget file computes it? You should be able to back-calculate that.

[TeunvanDooren]

It does give the same answer as the budget files. I checked it multiple times and there appeared to be a small mistake when referring to the connected cells in my code for output-processing. That's why I edited the post.
However, it is still relevant to know if there is a more direct way to obtain the MAW-concentration.

[christianlangevin]

It sounds like you are using the SSM Package to represent the effects of the MAW instead of using the MWT Package, which would explicitly simulate the concentration in the well itself. With the SSM Package any water that flows from the aquifer into the MAW, leaves at the concentration of the aquifer, whereas any water that flows into the aquifer from the MAW, comes in at the concentration specified in the SSM Package. So with the MAW/SSM approach you can just "observe" the concentrations in the model grid itself (which I see you have done), which should give you the information you need. You can "observe" a cell concentration with the GWT OBS Package. With the SSM approach, MODFLOW never calculates the concentration in the well, just the mass fluxes to/from the well so there is no way to observe it.

Alternatively, you can simulate the concentration in the well using the GWT MWT Package. The MWT Package will do a mixing calculation in the well bore and calculate a concentration. This mixing calculation may or may not be appropriate for you problem; with a freshwater/saltwater model, it may not be. With MWT you can observe the calculated well concentration using the observation utility.